The application of PI control and Smith predictors in an air-handling plant

This paper describes the application of PI control and Smith predictors in an air-handling plant. Results show that the PI control performance can be improved by introducing a Smith predictor with carefully selected parameters. The differences in set-point and disturbance response of these controllers are investigated. The results show that the introduction of a set-point filter can reduce the maximum overshoot in set-point response without having to decrease the gain of the controller. The results also demonstrate some of the typical nonlinear behaviour of air-handling systems. The cause of this nonlinear behaviour is analysed and some possible solutions are suggested